1. Field of the Invention
The present invention relates in general to a power interface circuit which controls the supply of power from a power supply to plant process equipment which is normally de-energized, and more specifically to a power interface circuit which controls pumps, valves, etc. in a pressurized light water nuclear power system.
2. Description of the Related Art
Plant process equipment, such as the equipment in a nuclear power system is conventionally controlled by redundant actuation circuits which receive input signals from redundant sensors. An example of a conventional two out of four voting power interface circuit for both normally energized and de-energized circuits is illustrated in FIG. 1. An example of how such redundant circuits can be used for two out of three voting is described in U.S. Pat. No. 4,200,864 issued to Gillet et al. In Gillet et al., two out of three voting is performed using relays; however, two out of three voting is also possible with valves, as described in U.S. Pat. No. 4,184,514 issued to Ryan et al.
In addition to the redundancy provided by conventional control circuits for critical plant processing equipment, such control circuits are conventionally designed to be testable to further improve their reliability. Usually, such testing requires the manual intervention of an operator with the result that the test of a single logic circuit necessitates several hours according to Gillet et al. Elimination of such operator controlled testing has been suggested for units such as line printers, see U.S. Pat. No. 4,317,412 issued to Bolcavage et al., and communication networks, see U.S. Pat. No. 4,320,508 to Takezoe. However, with respect to power interface circuits for plant process equipment, conventional systems are limited to self-testing of the status of contacts as described in U.S. Pat. No. 4,272,725 issued to Weiner et al. and do not include testing the circuit operation.
At least two problems must be overcome in order to produce power interface circuits which include automatic testing capabilities. The first problem area involves the complexity of redundant circuits. In order to replace the manual operations described in Gillet et al., complex automatic testing has been suggested, such as the multiplexer testing described in U.S. Pat. No. 4,422,140 issued to Keats.
The second problem area involves a conflict between testing the operation of an actuator and preventing inadvertent actuation of the device controlled by the actuator. Many of the systems in a nuclear power system are of the "fail-safe" type which are designed to be operated only in an emergency situation when another (independent) system fails to operate properly. Such "fail-safe" systems are required to be highly reliable and therefore, are preferrably testable. However, inadvertant actuation of such systems may cause partial or total shutdown of the nuclear reactor and require a re-start period of as much as several weeks or more. The conventional test circuit disclosed in U.S. Pat. No. 3,437,556 to Bevilacqua et al. is an example of a conventional system which uses relays and has to compromise between full testing of the system and the prevention of inadvertant actuations.